France

We will make boxplots of $\lambda$ ten days before and after the confinement begin from the results coming from the use of the different data sets (Johns Hopkins University for new cases and data.gouv.fr for new hospitalisations, see France_JHU and France_hospitalisations scripts). The data that led to the inference results used in this script is in both cases the data available 08/05 before 19h00.

In [1]:
import sys
import os
from importlib import reload

try:
    import covid19_inference.data_collection as cov19dc
    import covid19_inference.multi_country_plotting as cov19mcp
except:
    sys.path.append("..")
    import covid19_inference.data_collection as cov19dc
    import covid19_inference.multi_country_plotting as cov19mcp
WARNING  [theano.tensor.blas] Using NumPy C-API based implementation for BLAS functions.
In [2]:
reload(cov19mcp)
cov19mcp.boxplots_around_lockdown_one_country("France", distance=10)
The whiskers show 0.05 and 0.95 quantiles, outliers disabled

Italy

Original data from JHU (10 mai 2020) and Italian government GitHub repository (09 mai 2020).

In [3]:
cov19mcp.boxplots_around_lockdown_one_country("Italy", distance=10)
The whiskers show 0.05 and 0.95 quantiles, outliers disabled
In [4]:
reload(cov19mcp)
cov19mcp.plotly_violins_all_available_lambda(distances=10)

Violin plots of $\lambda -\mu$ for each of the countries / datasets found

10 days before and 10 days after lockdown

Country Dataset lockdown date lockdown type data retrieved
Netherlands JHU 23/03/2020 social distancing measures
Sweden JHU 24/03/2020 restaurant closures 15/05/2020 10:56
France new_hosp 17/03/2020 confinement
France JHU 17/03/2020 confinement
Italy new_gvt 10/03/2020 confinement
Italy JHU 10/03/2020 confinement
Abbreviation Meaning
JHU Data from Johns Hopkins university
new_gvt New cases according to government data of the respective country
hosp New hospitalisations according to government data of the respective country

For the links to the sources, see the readme of the repository.

Data has been pre-processed to include only data between 0.05 and 0.95 quantiles. The following plots are interactive, by clicking on a legend item, the corresponding plot will disappear. Double-click to isolate.

In [ ]: